A comparative study on techniques used for prediction of stock market

The most significant issue in the recent times in finance is finding the systematic ways to abridge and envision the stock market data. Stock market analysis gives useful information to Entrepreneurs, Individuals and Institutions about the etiquette of the market helping with speculation decisions. Many prediction models have been developed during the last decade. As a comparative study, we will be analyzing the ARIMA model, Neural Nets, Holt winters and the TSLM linear model. These models are outlined to assist the investors, bankers and capitalists to find out the concealed patterns that are present in the historical data. For the analysis purpose, we will be using five different companies' data procured through the Yahoo Finance. As the prediction of stock market is always considered as a provocative job and the precision of the predictions done has become an important part a comparative study is even done while removing the seasonal trends in the data too. We have observed that the Holt Winters model while removing the seasonal trends outperformed when compared to the rest models with a higher prediction accuracy.